Posts Tagged ‘Information Architecture’

 Tunard Garden

To plant is but a part of landscape composition; to co-ordinate is all.
Christopher Tunnard


My youngest sister is a freelance landscape architect in London, and so I know a little about this area.

As such, I have heard of Christopher Tunnard, a Canadian-born architect and academic, who in the early part of his career practised as a landscape architect. In 1938 he published Gardens in the Modern Landscape which married the influential modernist ideas then current within architecture with the discipline of landscaping.

Working in Britain at the time, the effect of Tunnard’s writings were short lived in Europe as the coming of the Second World War fostered a move towards more socially responsible design. Yet in 1939 he moved to Harvard where his theories later became the catalyst for what can be termed as the Anglo-American modernist movement in landscape architecture.

And in viewing a garden in the United States that showed his ideas in practice – sadly, the only surviving example of his landscaping work in America – I was struck by how similar planning in landscape architecture is to the visual processes that we employ within web development. 

Tunnard Garden - design

Tunnard's Garden: design plan










Site Mapping and Concept Modelling for Information Architecture

While anyone working in web development knows about the composition, usage and value of site maps, by contrast the related use of concept modelling is a relatively new and unexplored discipline (as applied to IA).

Basically, concept modelling is about conceiving and presenting an abstract representation of the informational ideas and their relationships found at any level within a site. [1]

As Dan Brown in his excellent book, Communicating Design (Peachpit Press, 2007), explains: concept models illustrate how different ideas relate to one another, representing the building blocks of the idea as nodes and their relationships as lines between them.

The diagram below from the book shows a concept model for a website selling musical instruments.

Concept Map - Muical Instruments

© Dan Brown, Communicating Design (2007)














As Brown says:

While this concept model doesn’t show any more of less detail than the others, it does deal with a specific concept. Instead of representing a broad view of the website, it shows all the different kinds of data on the site and how they relate to each other. This concept model illustrates the different ways of categorizing an instrument (the categories of categories) and associated metadata for each instrument. This concept model represents the relative importance of each concept by varying the size of the circles.

And the principal value of such concept models lies in their flexibility as a planning tool to tease out the underlying relationships between nodes.

Essentially they enable people to visualise (and then discuss) the multiple and complex relationships found between users, nodes (as areas, content types, and metadata, etc.) and the contextual links found in a complex system. 

Thus, they can be used in formulating an approach to designing a site structure, and should be used early within an information architecture process.

By contrast, a site map is more focused on defining the eventual structure of the information found in a website, and illustrate a part-whole relationship where an item lower in the map belongs to a higher level item.

And although they can be very different in presentation the traditional site map often looks very much like an org chart: a system of boxes representing pages, connected by lines representing links with their placement and linkages representing the hierarchy. 

Site Map

© Dan Brown, Communicating Design (2007)

Thus, a site map more often tends to the presentation of the end result of an information architecture (often expressed as a single A4 page in a hierarchy represented by uniform squares and links in tiered structure).

By contrast, concept models offer a flexible and dynamic approach to site map planning and may be used as a tool to assist in the construction of an information architecture, which will probably be turned into a site map for the final solution design.

Therefore, in effect, site maps represent the visual specification of a site, and are often used as part of a solution design document. While concepts maps are a tool that assist in the generation of such an information architecture specification.


Visual Mapping in Practice

Just recently at Storm we have started to use concept modelling in our own SharePoint projects.

Their value is of especial importance when one is having to consider the migration and re-design of an existing site structure in an attempt to improve the user journey.

Now this sort of “migration” scenario is a very common one in a SharePoint Intranet project, where an organisation might be considering one of two possible solution paths:

  • Version migration: A site migration from one version of SharePoint to another; or
  • Improving site structure within the same version: Examining and then improving a site structure – possibly badly conceived in the first place – within the same SharePoint version.

At Storm we are working on two such projects at the moment.

And the one that I have chosen to demonstrate the application of concept mapping is one that deals with version migration.

The project setting is a Scottish public body that has a large number of external stakeholders, for whom we are working on migrating (and improving) the design of an existing SharePoint Intranet while simultaneously planning the migration from SPS 2003 to SharePoint 2010.

Here the aim is to transition from a “file share” SharePoint approach that was adopted in SPS 2003 to delivering a higher-level Communications & Information Portal in SharePoint 2010, while simultaneously looking to use the features of 2010 to support and enhance the overall functions of the site.

As such, we are taking a large and relatively chaotic information domain (where all staff members may contribute) and seeking both to rationalise and re-organise it so that it becomes a more centralised and managed tool for information management for the organisation in question.


Project Process

Following James Robertson’s Enterprise IA methodology we started the design process for this particular project through generating an Intranet Development Roadmap (note: a 2 page word document – nothing more) and conducting Needs Analysis via a user survey (based on the Intranet Review Toolkit).

The objectives for the Intranet concept that emerged were presented as follows:

Intranet Development Roadmap


Next Steps  

With a clear and manageable objective in place and some early needs analysis of features to support, we are currently in the process of identifying user tasks/goals and starting an initial IA review.

And because there is already a SharePoint Intranet in existence, our early IA review is focused on the mapping this existing site (both structure and content), and generating an understanding of current logic behind the information organisation (with a view to revising this treatment).

For this we are using both site mapping and concept modelling (see high-level examples below).

Site Mapping

A site map of the top-level site structure for the existing Intranet reveals a fairly large information structure:

Intranet - Site Map

Intranet - Site Map

Concept Model

By contrast, the early concept model reveals something of the relationships we must explore further:

Intranet - Concept Model

Intranet - Concept Model

Well, what’s the outcome so far?

What is interesting is that by using this approach within the project team – rather than merely an interview-based methodology – we are beginning to understand how a potential IA restructuring might be of value across the Intranet.

Take, for example, the Corporate Reference Library – which for the purposes of this study we will call the “Knowledge Centre”. [2]

As currently conceived, this Intranet area is intended as a central informational repository for all staff to access key documents and external resources.  

As such, it has two main internal user audiences:

  • Staff Members in general
  • “HelpLine” Advisors: These represent a sub-section of staff who are dedicated to providing an advisory service to the organisation key external users (i.e. young adults).

As originally conceived, the Knowledge Centre was focused on catering for both user groups, aiming to deploy both key organisational documents and links to external key policy and informational resources organised by a generic sub-site hierarchy of “Subject Focus”.

See, for example, the sub-site structure for “Arts & Culture” below (which has the generic sub-site structural treatment presented below):

Knowledge Centre - Arts & Culture Site Map

Knowledge Centre - Arts & Culture Site Map

However, it appears from our initial reviews that there are a number of issues with the Knowledge Centre in its current treatment:

  1. It is trying to cater for two user groups – who have different goals and tasks. In supporting two user groups is actually trying to support two functions in one place: (a) a centralised informational resource for all staff and (b) a helpline resource for information advisors. As such, it does neither well.
  2. The role of the Centre within the Intranet is not clearly defined. As such, because all staff may contribute to the Intranet, the Knowledge Centre is not the only place where staff may add relevant information resources within the site. Other high-level areas such as Projects, Support Services and Products and Services also contain key information resources of a topical nature.
  3. The Centre does not support a poly hierarchical view of content. The Knowledge Centre’s current sub-site structure means that a document that has content relevant to a number of Subject Topics appears in only one Subject Focus in the hierarchy. And without content tagging by Subject Focus and access to a scoped search at the Knowledge Centre-level this results in staff not finding relevant content unless they already know it exists and where it resides in the sub-site hierarchy.
  4. The “HelpLine Advisory Service” is not well catered for in the existing treatment. In fact, it appears that the Advisory Service’s information resources are scattered in a number of places throughout the existing Intranet site (e.g. Support Services > Enquiry Answering; Knowledge Centre > All Subject Topics; Projects > Information Advisory Service; Products & Services > Various.).
  5. Key Content Types are not defined for users. Again in relation to the HelpLine Advisory Service, specialist content types such as Factsheets on subject areas are not immediately visible to Advisors and instead are scattered across a fairly large “Subject” sub-site structure. As such, given the nature of their job, Advisors find it difficult to trace relevant information quickly and rapidly whilst on the phone, and instead resort to individualised information strategies to answer enquiries.  

Early Conclusions

It appears from our initial mapping work – to be confirmed by actually interviewing the HelpLine Advisors – is that there is a distinct case to separate out the current treatment of the Knowledge Centre into two new and separate high-level site areas for the Intranet:

  • A HelpLine Advisory Service for Advisors
  • A Corporate Reference Library for Staff

As such, this treatment has been presented in an early concept model as follows:

HelpLine & Reference Library - Concept Map

HelpLine & Reference Library - Concept Map

And while we have not yet discussed:

  • The Content Types needed to support the (a) HelpLine Advisory Service and (b) the Reference Library areas; and
  • The logical organisation and structuring of each area – whether by sub-site structure and/or by metadata and scoped search.

Nevertheless, we have gained an important understanding at an early stage of what is deemed a key support area for the organisation through the use of visual mapping of an existing Intranet structure.

And Finally ….

I think there is much to be gained by adopting a visual mapping approach to the design of SharePoint sites; especially so with Intranets – which can be very large information structures that, by definition, often present complex information architectures to their users.

As is the nature of an Intranet solution, very often such a site goes through a series of site architecture revisions as it evolves in relation to the organisation in question. And arguably this applies as much to a SharePoint Intranet as any other solution; perhaps more so, because early information architectures for SharePoint have, more often than not, been badly conceived simply because the world over we are learning “best practices” with this product as we go.

In addition, the task of site migration from one SharePoint version to another is a key issue facing organisations in the decision to upgrade. As such, it is rare that an organisation will not want to take the chance to review its current SharePoint site treatment in an effort to improve information management support for its staff.

And clearly while information architecture is only one consideration in site migration within SharePoint, it is nevertheless a critical one and as a result we need to improve our conceptualisation of such architectures and our project approach to conducting this task.

I think that adopting a visual mapping approach to information architecture is of great value in SharePoint solution design. And, in particular, concept modelling is an important means whereby we can qualify and discuss our understanding of Intranet site maps within project teams in such a way that it improves our overall approach to information architecture.

Alongside more traditional tools such as Site Maps and Content Inventories it is yet another tool to assist in our ambition to design improved Intranets for our users.


[1] In Communicating Design Dan Brown summarises Concept Models as follows:

Concept Models - Dan Brown























[2] Note: in the interest of preserving client confidentiality, the references in this Case Study have been deliberately anonymised throughout this post.


Read Full Post »

Nick Drake

When I was younger, younger than before
I never saw the truth hanging from the door
And now I’m older see it face to face
And now I’m older gotta get up clean the place.

And I was green, greener than the hill
Where the flowers grew and the sun shone still
Now I’m darker than the deepest sea
Just hand me down, give me a place to be.

A Place
to Be
Nick Drake (1948-1974)
English singer songwriter of truly magical ability


As we all know, Australia is a place to be, and equally this seems to hold true in their information management practices.

I simply don’t know why this is the case historically, but nevertheless the fact is that this country has, quite disproportionate to its size, spawned a series of EDM systems, and their government consistently adopts an outstanding stance on information management.

Victoria Place to BeAnyway, the tradition continues to burn bright in the publication of Intranet Information Architecture: Best Practices Analysis (from the e-Government Resource Centre, Victoria State Government, Australia, December 2008).

And while this is focused on one particular Intranet implementation (and strictly speaking is not just about IA), the applicability and pedigree of this report makes it worth consideration, and includes findings and summations from the following stellar references in the web fields of usability, information architecture and intranets:

  •  Jakob Nielsen (User Advocate and principal of the Nielsen Norman Group)
  • Louis Rosenfeld (independent information architecture consultant, and founder and publisher of Rosenfeld Media, a publishing house focused on user experience books)
  • Gerry McGovern (widely regarded as the number one worldwide authority on managing web content as a business asset) and
  • James Robertson (Managing Director of Step Two Designs).

And while not specifically focused on technology, this best practice report is absolutely worth your time if you want to use SharePoint internally.

It is not phenomenally detailed, but in my book that is a virtue.

In my ongoing work with SharePoint I still see so many project teams struggling to provide clarity even on the basics in their own Intranet solution design, and yet this short read suggests some of the key – platform agnostic – practices that you need to think about in implementing an Intranet on SharePoint.

Intranet Information Architecture Best Practices - excerpt

Twenty-three pages later and maybe you’ll see some “truth hanging from the door….”

Read Full Post »

As someone that likes his books, yesterday I discovered that a new book entirely devoted to solution planning using MOSS 2007 was on the cards.

MS SharePoint Planning, Information Architecture & Design Bible

MS SharePoint Planning, Information Architecture & Design Bible

This is probably very old news for those interested in this specific area, but nevertheless I am very interested in the appearance of this book.

It’s indicative of the growth of the global industry spawned by SharePoint that this specialist title is on the way at all (all 792 pages of it). As an ex-Commissioning Editor myself, I know that Publishers simply wouldn’t be bothered with this book if they didn’t think there wasn’t a viable market.

There is, for instance, very little risk in publishing SharePoint for Dummies (released 5 months after MOSS 2007 RTM-ed in late October 2006) and clearly no-one in publishing is going to get sacked for making that call.  However, it is quite another decision to publish a book that is for a specialist, and somewhat non-technical, sub-audience of the platform 2+ years later. It shows one just how much the MOSS market has grown in the interim period, and where some of the attention in implementing ECM solutions on SharePoint is now being focused. 

And a quick search on Google for “SharePoint Information Architecture jobs” reveals just how much SP Products and Technologies is starting to dominate – rightly or wrongly – the ECM industry. When a product starts to generate this sort of specialist role requirement it’s yet more evidence of a healthy future.

I don’t know David Sterling’s work – as an MVP he has published books on MOSS previously. And while I didn’t like the more general Bible title on MOSS  – the word “bible” says it all, really – and so I gave up on it quickly, the one above has a guaranteed book sale of at least 1 in the UK when it RTMs.

Read Full Post »

As one of the softer web disciplines that lies at the heart of Microsoft SharePoint as a software product and its central importance in the implementation of countless solutions, I have read a lot of articles and book chapters about SharePoint that stress the importance of “getting your information architecture right”.

The quality of this thinking is not always quite what it might be. The analysts at Forrester, for instance, tell us (seemingly with great authority):

SharePoint buyers expect intuitive navigation, contextual search, and easy administration out of the box. But such benefits depend on how content is structured, labeled, and categorized, and they require a nuanced understanding of how different audiences will navigate and search for information.

The information architecture (IA) behind a SharePoint deployment has lasting consequences for the end user experience and for Web site management. Information and knowledge management (I&KM) professionals should use their SharePoint implementations as an opportunity to set solid information architecture in place that turns today’s information overload into tomorrow’s valuable information assets.

The upshot?

Information workers will finally be able to find the critical information they need to do their jobs.

(See Forrester: The Critical Role of SharePoint Information Architecture, http://blogs.msdn.com/architectsrule/archive/2009/02/26/forrester-the-critical-role-of-sharepoint-information-architecture.aspx)

Ok … so far, so good, but where exactly does this get us?

Town Planning… 

Town planning

Town Planning (© things magazine)

In early 1998 Louis Rosenthal and Peter Morville made their name in publishing a bestselling book entitled Information Architecture for the World Wide Web (now in its Third Edition!).

As a newly appointed e-business consultant working for The iGroup at Computercenter in London, I remember avidly reading their – essentially very simple – book and thinking “Wow, these guy’s have got it right!”

Switch to mid 2009 and inevitably the practice of building websites is far more sophisticated, and hopefully I am little less naïve than I used to be.

Superficially, at least, it seems that those professionals who devote their lives to working on the web are now much more qualified to provide a sophisticated response to the question of what makes a “perfect” information architecture. One that takes into account so-called “best practice” in emergent web disciplines as diverse as needs analysis, stakeholder and user audience profiling, content analysis, information science and classification, usability and accessibility, wire-framing, web and user interface design.   

Yet growing sophistication in the field comes with a price.

Quite simply, it is very hard for a busy web professional – and I include myself in this category – to have a genuine appreciation of all these burgeoning web disciplines to allow true mastery over the undoubted “art” of information architecture. 

In my personal view, therefore, the vast majority of those who construct information architectures for a living are informed not by a deep understanding (of a myriad of web sub-disciplines), but rather by a mix of unvalidated opinions, their personal preferences, and a partial understanding of their trade.


In the City of Organised Thought

What of the subject matter of the information architects – the  websites, extranets and intranets?

Essentially, all these forms of “web patterns” reflect human activity in all of its diversity. This relationship is undoubtedly an organic thing that both evolves overtime and exhibits great complexity.

Think, for instance, of an Intranet that must by necessity reflect an organisation’s ever changing social, political and informational life, and you realise that this is indeed a complex undertaking. Likewise, a website that marries a particular content focus with this or that user audience – the latter always demanding, whoever they are.  

As such, these generic “web patterns” are capable of order and structure – indeed in many ways they exhibit and need such order – but the kinds of order that are possible vary enormously even within the same general setting.

Sure, these different kinds of order may be related as “patterns”, but equally they cannot be mutually reduced except by gross simplification.

Recognition of this awkward fact – i.e. a singular lack of neatly tended streets in our metaphor – also tends to be fiercely resisted by professionals who are engaged in theoretical system building both to justify their profession and the need to prove that they “know their stuff”.

And because they have succeeded in introducing an impressive kind of order they are always tempted to extend it more widely than they should. In effect, they have formulated a neat and tidy “theory” and as such want to place that above all others.

MEL Generated Cityscape by CGZool

Cityscape (© CGZool)

It is as though they have the ambition of knocking down the irregular and awkward city streets that disrupt their thinking, and rebuilding them to their vision – with a strong preference for unfailing regularity in both architectural design and pathways. 


Beware the Despotic Pattern with SharePoint

What then are some general pitfalls in town planning with SharePoint Products and Technologies?

First, there is the ever present tension of dealing with a software product in solution design. It is obvious that the very fact that SharePoint is a product both structures and constrains your thinking in relation to the practice of information architecture.

And while there is little doubt that MOSS 2007 as a CMS has an elegant hierarchical model enshrined in its Content Database structure:   

  • Site Collection
    • Sites
      • Content Types
        • Page Layouts
          • Controls  
          • Site Columns
            • Content
            • Metadata

Yet MOSS is known to be hard to customise….

As a consequence this model tends to impart a strong solution preference in formulating an IA within SharePoint. Effectively, out of the box MOSS 2007 – in either its Publishing or Collaboration Portal mode –  will do what it wants to do very well indeed; but it is not quite so good when your architectural design does not conform to SharePoint’s particular blueprint.

There is, therefore, no doubt that it is very easy to lose honesty very quickly in the overwhelming desire to avoid the pain of customising SharePoint to meet a requirement for implementing even a moderately awkward neighbourhood.

Try, however, to resist the lure of this particular siren!

As we have discovered at Storm just recently there are some very real benefits in remaining honest. This being due in the main to the technical brilliance of my colleagues – in particular Tom Travers – who have succeeded in opening up MOSS 2007 to customisation in a number of different areas. Town planning being one of them.

Second, beware Greeks that bear false gifts… and retain a healthy skepticism towards the town planners of this particular product (i.e. the SharePoint analysts that solemnly tell you how to formulate that perfect IA). Telling is one thing; doing is quite another. Thus, treat with caution – though clearly don’t dismiss – their formulas and pronouncements as they relate to the complex and difficult task of town planning.

And, finally, never lose sight of the fact that in constructing an IA – regardless of whether it is for SharePoint or not – there is no substitute for deep understanding, imagination, and, yes, a lot of hard work.

Note: This post is the first of an occasional series that attempts to explore the discipline of information architecture in relation to SharePoint. And my advance apologies if this first post mentions relatively little about SharePoint! I am simply trying to explain how I see the landscape of the discipline before turning my attention to SharePoint proper.

Read Full Post »